MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641048373 A) filed by Divya S; Kanish K; Kathiravan G; Lehan Zidin; Rakesh TS; Visveshwaran DV; Shravan Kumar B; Tharun Aadhith MS; and Raja S, Coimbatore, Tamil Nadu, on April 16, for 'deep learning-based leukemia classification system.'
Inventor(s) include Divya S; Kanish K; Kathiravan G; Lehan Zidin; Rakesh TS; Visveshwaran DV; Shravan Kumar B; Tharun Aadhith MS; and Raja S.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a deep learning-based leukemia classification system designed to provide an automated, accurate, and efficient approach for the detection and classification of leukemia using microscopic blood smear images. The system integrates advanced artificial intelligence techniques, particularly convolutional neural networks (CNNs) and transfer learning models such as ResNet, to analyze complex morphological features of white blood cells and distinguish between normal and leukemic samples. The proposed system comprises multiple functional modules, including data acquisition, preprocessing, segmentation, model core, post-processing, and monitoring. Input data, consisting of raw datasets and newly uploaded images, are stored with associated metadata for traceability. The preprocessing module enhances image quality through normalization, noise reduction, and contrast adjustment, followed by segmentation and region of interest (ROI) extraction to isolate relevant cellular structures. The model core performs feature extraction and classification using deep learning techniques, enabling both binary and multi-class classification of leukemia types, including acute and chronic variants. The post-processing module calculates clinically significant parameters such as blast cell count and percentage levels, and generates alerts based on predefined thresholds, assisting in severity assessment and triage. The system also includes a user-friendly interface for clinician interaction and monitoring modules for continuous evaluation of performance. The invention provides a scalable and robust solution that improves diagnostic accuracy, reduces manual effort, and supports timely clinical decision-making, making it suitable for deployment in diverse healthcare environments."
Disclaimer: Curated by HT Syndication.
